The Exceedance Design Effect: Effective Sample Size for Thresholds under Clustering
Many machine-learning systems set a threshold at a quantile of a calibration set: conformal predictors that promise 90% coverage by drawing their cutoff at the calibration set's 90th percentile, abstention gates that decline to answer when a model's score falls below the calibration set's tenth percentile, safety filte...
A. Noonan
· 0 citations